Self-contained minimum example script to run scrapy

runner.py

import json

from scrapy.crawler import Crawler
from scrapy.contrib.loader import ItemLoader
from scrapy.contrib.loader.processor import Join, MapCompose, TakeFirst
from scrapy import log, signals, Spider, Item, Field
from scrapy.settings import Settings
from twisted.internet import reactor


# define an item class
class DmozItem(Item):
    title = Field()
    link = Field()
    desc = Field()


# define an item loader with input and output processors
class DmozItemLoader(ItemLoader):
    default_input_processor = MapCompose(unicode.strip)
    default_output_processor = TakeFirst()

    desc_out = Join()


# define a pipeline
class JsonWriterPipeline(object):
    def __init__(self):
        self.file = open('items.jl', 'wb')

    def process_item(self, item, spider):
        line = json.dumps(dict(item)) + "\n"
        self.file.write(line)
        return item


# define a spider
class DmozSpider(Spider):
    name = "dmoz"
    allowed_domains = ["dmoz.org"]
    start_urls = [
        "http://www.dmoz.org/Computers/Programming/Languages/Python/Books/",
        "http://www.dmoz.org/Computers/Programming/Languages/Python/Resources/"
    ]

    def parse(self, response):
        for sel in response.xpath('//ul/li'):
            loader = DmozItemLoader(DmozItem(), selector=sel, response=response)
            loader.add_xpath('title', 'a/text()')
            loader.add_xpath('link', 'a/@href')
            loader.add_xpath('desc', 'text()')
            yield loader.load_item()


# callback fired when the spider is closed
def callback(spider, reason):
    stats = spider.crawler.stats.get_stats()  # collect/log stats?

    # stop the reactor
    reactor.stop()


# instantiate settings and provide a custom configuration
settings = Settings()
settings.set('ITEM_PIPELINES', {
    '__main__.JsonWriterPipeline': 100
})

# instantiate a crawler passing in settings
crawler = Crawler(settings)

# instantiate a spider
spider = DmozSpider()

# configure signals
crawler.signals.connect(callback, signal=signals.spider_closed)

# configure and start the crawler
crawler.configure()
crawler.crawl(spider)
crawler.start()

# start logging
log.start()

# start the reactor (blocks execution)
reactor.run()

scrapy.contrib is deprecated, they recommend to use this one instead:

from scrapy.loader import ItemLoader
from scrapy.loader.processors import TakeFirst, MapCompose, Join

You could use this also:
process = CrawlerProcess({ 'ITEM_PIPELINES', { '__main__.JsonWriterPipeline': 100} })
process.crawl(DmozSpider)
process.start()

No need for a callback or starting/stopping the reactor manually

This was exactly what I was looking for!!! I've been wanting to transition my Scrapy projects to stand-alone for a while now. Thanks!!!
Here's how I ended up doing the settings:

settings = Settings({
    # piplines start with the project/module name so replace with __main__
    'ITEM_PIPELINES': {
        '__main__.WriterPipeline': 100,
    },
    'DOWNLOADER_MIDDLEWARES': {
        'scrapy.downloadermiddlewares.useragent.UserAgentMiddleware': None,
    }
})
process = CrawlerProcess(settings)

# you can run 30 of these at once if you want, e.g —
# process.crawl(CustomSpider)
# process.crawl(CustomSpider) etc.. * 30
process.crawl(CustomSpider)
process.start()